Nucleic acids are under-represented molecular targets for small molecule therapeutic agents, even though recent research shows that unusual nucleic acid structures play a profound role in gene expression and DNA replication. We propose an innovative, interdiscplinary, high-throughput screening approach for the discovery of new nucleic acid-targeted therapeutic agents. Our approach integrates an actual experimental screening assay with a virtual screening assay. The former is a unique competition dialysis assay that we invented for the discovery of small molecules that target specific nucleic acid sequences or structures. In the competiton dialysis assay, an array of designed nucleic acid structures and sequences are dialyzed against a common test ligand solution. At equilibirum, more ligand accumulates in the dialysis cell that contains the preferred structure or sequence. The assay is thermodynamically rigorous, and allows for quantitative measurement of ligand binding free energies. In our integrated approach, an in silico virtual screening assay will be implemented that will contain an array of receptors that is identical to the actual array of nucleic acids used in the competiton dialysis assay. The virtual assay can easily screen millions of compounds for their selective binding to particular nucleic acid structures of functional significance. "Hits" from the virtual screen will be passed to the actual competition dialysis assay for validation. With proposed improvements, the competiton dialysis can rapidly screen hundreds of compounds. Our strategy will be to identify libraries of compounds that selectively recognize particular structures with the virtual screen, then to rigorously verify these "hits" by high-throughput competition dialysis. Specific plans for the development of the competition dialysis assay include : i) implementation of the assay in a 96-well plate format (facilitating automation), ii) reduction of sample volumes, iii) expansion of the biologically significant nucleic acid array, and iv) reduction of throughput time. The virtual and actual screens will be fully integrated by an iterative feedback loop to refine the scoring algorithm, using actual data obtained from the first generation dialysis assay in which the interaction of 126 compounds with 13 nucleic acids structures was studied. Our approach matches the Road Map Initiative goals by its interdisciplinary approach and by its emphasis on expanding the toolbox available for the discovery of new therapeutics.